Questions, Answered
Frequently asked questions
What is SAGE x PM?
The project-management side of SAGE, CinePipeAi's AI Creative Director. It's how we plan, prioritize, and track the work of building and running CinePipeAi, using the same agent-orchestration thinking that already runs our production pipeline.
Is this the same SAGE that directs creative decisions in the Creator app?
Same engine, different job. In Creator, SAGE recommends agents and creative direction for a production. In PM, it's turned toward the roadmap itself: the tasks, goals, and agents that build and run the platform.
Who is this for?
Every team building on CinePipeAi's platform, from our own internal roadmap to any studio running SAGE-orchestrated production work. It's the same tool whether you're running on our infrastructure or your own.
What does "Require SSO" actually do, and what happens if my company turns it on?
"Require SSO" is a per-domain toggle an org owner turns on after verifying a company email domain (Settings -> Domains & single sign-on). Once it's on, signing in with a personal Google/GitHub account under that email domain is blocked outright, and it also immediately signs out everyone on that domain already logged in with a non-SSO session, not just future sign-ins. It's a deliberate security choice, the same way tools like Slack or Notion let a company lock down sign-in to their own identity provider. Turn it on only once you have another confirmed way in: full SSO federation with your own identity provider (Okta, Azure AD, etc.) isn't supported yet, so today this only blocks casual Google/GitHub sign-in for your domain.
Why doesn't this look like a sprint board?
Because the work isn't sprint-shaped. Sprints assume a fixed two-week commitment and a person re-planning it by hand. Here, work is a dependency graph, so re-scoping is an edit to a task, not a decision about which sprint it lives in, and the system stays continuously ready rather than waiting for a planning ceremony.
What's a "wave"?
A wave is the batch of tasks the system dispatches together right now, because they're unblocked and ready. It's not a sprint or a deadline; a wave can be as small as everything that just became ready this instant.
What decides what to work on next?
A continuously recomputed ready-queue: everything unblocked, ranked by how much it matters against real goals, not a backlog someone re-argued in a meeting.
What is "confidence" tracking, and why is it separate from status?
A task can be "in progress" and still be poorly understood. Confidence is tracked as its own signal, built from open questions and how sure the work actually is, so "we don't know how to do this yet" is something you can see, not something quietly absorbed into a due date.